This set was a gift from a neighbour of my parents. OK, maybe the word "gift" gives the wrong impression - for this radio it was either me or the rubbish bin!
It had been stored in a damp garage or shed for several years. The case was falling to pieces, and one corner showed signs of woodworm and rot. As soon as I received it, I wrapped it securely in polythene bin bags and left it in the garage. I wasn't letting it into the house until the woodworm was treated, and that would have to wait until the weather was reasonable.
Most of the veneer on the front had come away, but the person who rescued the set had retrieved all the pieces she could find and put them in a plastic bag. The tuning scale was broken, but fortunately near one end where there are not many markings. Apart from that, the set appeared to be complete and showed no obvious signs of previous repairs.
The photo on the right shows the set as I received it - although it looks better in the photo than it really was! There is a photo of the finished set at the bottom of this page.
When the weather was better the set was unwrapped from the polythene - still in the garage. The top and two sides of the case were loose and were just lifted away, as was the back. I then set about removing the chassis from the front/base assembly. The magic eye and speaker were held in place with screws, which came out with no problems.
It is interesting to note that the top and bottom edges of the speaker had been cut away by about ¼" each, so it would fit into the cabinet. I initially thought it was a replacement speaker, but there were no signs of previous fixing holes and the soldering looked original. The cutting away was clean and tidy - probably too good to have been done as a repair, so I concluded that this must have been done in the factory.
The knobs were fitted using grub-screws. Three came out OK, but the fourth would not budge. I applied some WD40 and left it for a few minutes, but this didn't help.
I tipped the set onto one end and removed the access plate from the base. Fortunately the jammed knob was on the tuning control, and by releasing the screw holding the flywheel onto the shaft, the shaft could be released. I had to loosen the tuning drive cord by releasing the spring from the wheel on the tuning capacitor, and could then remove the knob and shaft as one assembly. This would, of course, have to be sorted out properly before the set is reassembled. But for now it allowed me to continue.
The chassis was fixed to the base with four large screws, which came out with no problems.
The woodworm holes seemed to be confined mainly to one corner of the cabinet. There was no sign of new woodworm activity, even though the set had been in the warm garage for most of the summer. However I was taking no chances!
I sprayed a generous quantity of Cuprinol Woodwork Killer (the type that comes in an aerosol can with a pointed nozzle) into every flight hole I could find. As well as the main areas on the side and base, there were a couple of other holes in other panels. Although they were probably not woodworm holes, they were treated anyway. The whole job used over half a tin of the woodworm killer.
The pieces of the cabinet were then left in the garage (on polythene on the bench) for several days to dry out.
Chassis Initial Checks
Before going too much further, I decided to give the chassis a quick check over for any major (expensive) problems that could make the whole job not worthwhile.
Initial resistance checks on the mains and output transformer showed no problems. I connected the capacitor reformer to the main smoothing capacitor, which came up OK after a few minutes.
Time for the real test. I connected the speaker and an aerial, then connected the chassis to the mains via the lamp limiter. A test meter across the HT showed nothing unusual, and after the usual warm-up delay I was rewarded with static and crackles. Operating the waveband switch and the tuning capacitor bought in a few stations. It was clear that the set worked fairly well on MW and LW, but not at all on VHF. This is probably the reason the original owner stopped using it years ago. A quick voltage check on the grid of the output valve showed the usual few volts positive due to the leaky coupling capacitor. I cut one lead and temporarily soldered a modern replacement in its place, which solved that problem.
This was sufficient to tell me there were no major problems. I will probably need one or two new valves for the VHF tuner, and it could also do with a new magic eye (the existing one is very dim).
Several months later (longer than planned, but other jobs and a new CD-ROM got in the way) I returned to the cabinet. The woodworm killer had dried completely and, although the weather had been warm, there were still no signs of woodworm activity. I brushed off the worst of the dust, and bought the pieces into the workshop.
Before I could start reassembling the panels I had to refix the sections where the plys of the plywood were coming apart. This was done by the usual method of squeezing in Evo-Stik woodworking adhesive, pushing it into the gaps as far as possible with thin card. I then clamping the layers together between pieces of wood (using polythene between the cabinet and clamp pieces to prevent sticking), and used either clamps or heavy objects to hold it while the glue sets.
There were a number of sections that required this treatment so it had to be done in several stages (I only have six clamps). Although the glue bottle states that it takes 24 hours to fully set, I found that after about two hours I could remove the clamping and the bond would stay in place on its own, so I could move on to another section. I also glued several small triangular corner pieces that would be inaccessible when the cabinet was reassembled back into place. It was not possible to clamp these so I simply pressed them into place and let gravity hold them there while the glue dried. This whole process took a day (on and off).
The two fabric-covered sections of the front panel are held into the frame with a number of screws. I considered removing these to make the job of refinishing the rest of the cabinet easier, but most of the screws would not budge so I decided against this idea. Also these panels seemed to be the only things holding the front frame together now the case was in pieces.
The following day I started on the actual reassembly. Because some of the panels were slightly warped it was clear from a dry-fit that I would not be able to secure the whole thing together in one go. I started by fixing the side panels to the base/front assembly using more Evo-Stik woodworking adhesive. This was held together with clamps on the front edges and a bungie strap (length of elastic with a hook at each end) right around the cabinet near the bottom to pull the back corners in to the base. To help keep the assembly square, the top was slotted approximately into place as a support. A couple of corners were still not sitting snugly but I couldn't find a way to hold them in place with the limited selection of clamps I have. Instead I used a couple of small panel pins, hammering the heads below the surface with a punch. I already had plenty of woodworm holes to fill later, so a couple of nail holes didn't matter!
After about three hours I released the clamps
etc., and glued the top panel into place. This was the most distorted and needed
four clamps along the front edge and a couple of heavy transformers to hold
the back corners. The bungie strap was used again to pull the sides in, and
a couple more panel pins held the awkward corners. I then checked for square
at the back, by measuring the diagonals and by checking the back fitted correctly.
Once this had been drying for a couple of hours or so I glued the corner support blocks back in place. They were light, and the wet glue was sufficient to hold them in place in the top corners. I then left the cabinet for several days to ensure the glue was fully set, before moving on to tidying it up.
Cabinet Repair and Refinishing
The "proper" way to approach a cabinet like this would be to replace all the rotten/damaged wood with new panels, apply new veneer if necessary, stain the new woods to match the original and then refinish the cabinet in a style matching the original appearance. The result would be a very nice set.
This would involve significant expense, and would also require carpentry tools (such a router), ability and experience that I don't have. If this were a classic set from the 1930s or a rare model from the 1920s, this sort of restoration would be viable, and indeed would be the only appropriate approach.
However this isn't such a set. It's a run-of-the-mill set from the mid-1950s, and examples with reasonable cabinets are still available at swapmeets for about £30. It therefore did not make sense to spend a fortune restoring my set, particularly since the result would probably still not be as good as an original. I decided that I would do the best I could with the materials to hand, and maybe spend up to £20 or so on materials if necessary. Fortunately I already had a reasonable accumulation of stains, wood fillers and polishes etc.
Repair and Filling
I filled the woodworm holes, nail holes, gaps in the joints and other damage with wood filler. The bottom front corner of the left panel was quite badly rotten. I removed the flaking pieces that were beyond hope, then started to rebuild the section using wood filler. This had to be built up in several layers because the filler will only dry to a depth of about 3mm. Once I had achieved about the correct shape it was shaped and flattened with a file.
The pieces of veneer that were still in place were re-glued (again with Evo-Stik woodworking glue) where they were lifting, and held with masking tape while the glue dried. Some of the loose pieces could be pieced into place, and the remaining bits were just sufficient to finish the top edge and two side edges. This gave consistent colour to these areas.
The bottom edge was veneered with some new veneer that I had left over from another job. It was a lighter colour but I hoped I could stain it to give a passable match. At least being a separate complete run it wouldn't matter so much if it didn't quite match. I also glued a piece of veneer over the wood filler repair on the side panel (again not an exact match but better than filler). The slight gaps where the veneer pieces fitted together were filled with wood filler.
The next job was to remove what remained of the original finish. This appeared to be a sprayed varnish, and much of it had already flaked away. Most of the remaining finish was scraped off using a sharp paint scraper. I found it most effective to scrape at about 45 degrees to the grain.
The varnish on one side was a bit more stubborn so I had to use some chemical stripper. Because I was doing the job indoors I wanted something that was not too smelly. I also wanted something fairly gentle since I am only trying to remove one thin coat of varnish, not several thick coats of gloss paint. I purchased a bottle of Nitromors "Super Strip" which claimed to be "safer to use" and "low odour". The instructions said that some paint removal may occur after 1-2 hours but for best results leave for 4-6 hours, so it is obviously a fairly gentle and slow-acting product. I found that it had softened the varnish sufficiently to allow it to be scrapped off after only 20 minutes, so this is a good choice of stripper for our purposes. It also has a brush in the cap that makes application easier, and it doesn't burn if you get it on your skin (as long as you wipe it off straight away). The stripped area was then cleaned with meths to remove any residual stripper.
The only section that was not stripped was the inside edge surrounding the front panels. The varnish here was in reasonably good condition. This section is a lighter colour than the rest of the cabinet, and I felt by leaving the varnish untouched it would act as a mask if I accidentally got wood-stain on that area. The rest of the cabinet was then lightly sanded with fine (120 grit) sandpaper.
Staining and Finishing
I tried the wood-stains I had on a piece of scrap plywood to get an idea of the colour. The best one I had was Indian Rosewood. This seemed to be dark enough to bring the various odd coloured sections together, although it was possibly a bit too red. I added some of this to a nearly empty tin of American Walnut (which on its own was too dark). This gave a combination that was not so red and also not too dark. I applied this with a rag as instructed, and left it over-night to dry before rubbing off the excess. The result was very pleasing. As I had hoped, the colour was reasonably consistent, and the sections that were obviously different previously now toned in very well.
Since the set originally had a varnished finish, I felt it was best to stay with this. Varnish also has the advantage that it will soak into the wood and give a bit more support to the slightly flaky plywood corners when it dries. Because the front sections were in place I felt spraying the varnish would not be a good idea because it would be difficult to mask the area completely. Also it was now approaching winter and it was too cold, damp and windy to be outside doing the job with aerosol cans.
I purchased a tub of Cuprinol Varnish for interior wood, in a satin finish. I selected the Antique Pine colour because this was about right for the unstained section and I thought the slight aged yellow effect might soften the new stain look a bit. The product claimed to be "low odour" which is an advantage when working indoors, and the brushes can be cleaned using water. I also bought two B&Q fine finish paint/varnish brushes (12mm and 25mm), because I wanted to try to avoid visible brushstrokes as much as possible. These items, together with the stripper mentioned earlier totalled to about £20, which was the figure I had budgeted for.
Although it gave a reasonable finish, I found the varnish tended to dry rather quickly. It was therefore necessary to apply it fairly quickly and to avoid overlapping any varnish that had been applied more than a minute or two ago. For this reason I would not recommend this product for refinishing valve radio cabinets. Something slower drying would be a better choice. However, after three coats (24 hours between them) I had achieved a satisfactory finish. The solution was to leave the window open, the door shut and the radiator off for 24 hours before applying the final coat, so the room was cooler, and the varnish dried slower.
I wrote to the manufacturers explaining what I was doing and the problem I had, and asked their advice on applying the product or for suggestions of alternatives. They sent back a standard apology letter and a £10 voucher, and did not answer any of my questions!
Once the varnish on the cabinet was fully dry, the brass strips were polished with Brasso wadding and the other fixed trim was cleaned.
While waiting for various glues, stains and varnishes on the cabinet to dry, I tackled some of the other trim. Part of the red enamel from the Murphy badge had chipped away. I removed the badge from the cabinet (it was a press fit) and filled the offending areas with an appropriate coloured enamel model paint. To achieve the correct shade I had to mix some red with a small amount of black and an even smaller amount of a reddish-brown. After a week (to ensure the enamel was fully set) the excess was cleaned away using Brasso wadding.
The two sections of the broken tuning scale were carefully cleaned on the outside only. I used foam cleaner (applied to the rag not to the glass) to remove the worst of the muck, then a paint scraper to remove the remains of Sellotape adhesive (presumably a previous attempt at repair) and other dried dirt. I then used Brasso wadding to clean the specks the paint scraper didn't get, then polished off with a soft duster. The edges of the crack were cleaned with isopropyl.
With the dial face up on a piece of tissue paper on the bench, I carefully positioned the two pieces together. A strip of masking tape was then placed along the join and pressed firmly down. I then turned the whole assembly over. By lifting the broken area the joint would open, and it would close correctly when laid flat. I opened the joint this way and applied some superglue along both edges (the version with a brush in the cap is the best for this job) then laid the scale flat and left it for half an hour. With the tape removed the result felt secure, although it obviously isn't very strong. The crack can be seen if you get the light at certain angles, as would be expected.
Loctite also produce a glass adhesive that has a refractive index the same as glass and claims to give invisible repairs, but this needs UV light or bright sunlight to cure. I didn't have any of this glue to hand, and I also don't have a UV lamp and there isn't much bright sunlight approaching winter!
The print was slightly chipped by the break, but since this was in an area of the dial that is mainly black it was an easy matter to paint this area black again using enamel model paint.
The scale was originally fitted into the cabinet by four metal brackets pressing onto small rubber pieces fitted onto the edges of the glass. The rubber pieces had become hard, so I used four rubber sleeves slit along their length. When I came to fixing it into place however, I found that the front of the cabinet was not flat. The scale was in contact with the cabinet in the centre but raised at the ends. If I were to use the original fixings I would stress the glass and probably break it. The break I had repaired was close to one pair of fixings. I assume that the cabinet was either warped when made or more likely became warped over time, stressing and eventually cracking the scale.
Clearly I couldn't use the original fixing method. Instead I needed something that would support the glass without applying any stress to it. With the cabinet laid face down I laid the scale into position and supported it with some pieces of cardboard so that it was slightly away from the cabinet in the middle and spaced away equally at both ends. I then fixed it into place with a few small dollops of hot-melt glue. In the area of the repaired crack I placed the glue dollops by the repair so that it bridged both pieces. With the glue set and the cardboard supports removed the scale felt secure. However if the cabinet should continue to warp significantly, the rubbery hot-melt glue will hopefully stretch and become unstuck from the wood before the glass breaks.
Despite soaking in WD40 for several days, I still could not shift the tuning knob grub screw. There was no option but to drill it out. I held the shaft securely in a vice having plastic jaws. Using a variable speed battery drill and a drill bit that was a comfortable fit in the grub screw hole, I slowly and carefully started to drill into the screw. I had to drill away almost the whole screw before the knob started to loosen on the shaft. I then pulled the knob off the shaft then wound the remaining small piece of the screw into the centre of the knob using the drill bit hand held. The piece could then be shaken out. The shaft survived completely unscathed. The knob also had sufficient thread remaining to allow a 2BA screw to be screwed in and able to grip the shaft. I made a grub-screw by cutting a short threaded section from the end of the 2BA screw and cutting a slot to take the screwdriver.
Despite operating the tuning capacitor on the set during testing, the drive cord was still largely in place. With the cord tension spring released I was able to loop three turns of cord around the shaft and fit it back into position. When the tension spring was refitted the drive worked fine, and the pointer landed within 1mm of the reference mark on the backplate at the end of the travel.
Several months ago I had concluded that there was nothing seriously wrong with the chassis, apart from the lack of life on VHF and the very dim tuning indicator. Before continuing with the repairs, I tested the chassis again, to make sure I was starting where I had left off. I also replaced the dial lamps, because these give a good visual indication that the set is on and live. I noticed that the voltage selector was set to 220-230V, so I moved it to 240-250V to suit our local mains voltage.
I had ordered new replacement valves for the VHF head (EC92 and 6L34), together with a so-so used 6M2 tuning indicator from Valve and Tube Supplies, so the first job was to try these and see if they made any difference. They didn't. The FM ratio discriminator detector diodes are a 6D2 valve, so I replaced this with a CV140. The heater did not light so I wiggled the valve a bit. This got the heater working and bought about some crackles. With the set switched off I applied some contact cleaner to the valve socket, then refitted the valve. VHF now worked, although the volume needed to be set at nearly maximum to get a reasonable sound level.
I tried the original valves again in turn. The original 6D2 worked but the sound was distorted. The original 6L34 was also OK, but the EC92 was dead. The EC92 is the FM oscillator so if this stops working there will be no reception, whereas the 6L34 is the RF amplifier so something should still get through even if it is fairly low emission. I left the three new valves in place.
The tuning indicator was lit (an improvement on the original!), but didn't move from the minimum position on any waveband.
A few voltage checks showed that the HT was only 145V - it should be at least 250V. The transformer output voltages were OK, so I suspected the rectifier valve. This is a UU9, so I tried an EZ40, which is a direct equivalent. The HT was then 275V, but this made no difference to the performance of the set. Nevertheless, the EZ40 remained fitted.
There was one previous repair evident. The section of the waveband switch that selects whether the HT is routed to the VHF tuner head or the AM mixer-oscillator valve anode (via 33K resistor) had presumably suffered from tracking, because the connections had been removed and linked together. This meant that the HT was applied to both circuits whatever waveband the set was switched to. The set seemed to work OK wired this way. I wondered whether this might have something to do with the lower audio level on VHF, but disconnecting the 33K resistor (so the circuit was as it would have been on VHF with the switch connected) made no difference. I decided to come back to this later.
Further checks showed that it was not actually very loud on the AM bands either - the volume control was above half-way for a reasonable listening volume. This had to be due to either the triode amplifier or the output pentode.
I replaced the 6LD3 double-diode-triode with a new EBC41 (direct equivalent), which was a vast improvement. Voltage checks around the output pentode (6P1) showed that this was also a bit down on emission (the cathode voltage should be 8V but was only 6V), however it was working well enough and I didn't have a suitable replacement to try, so I left it.
So that's four valves needed (five actually replaced), plus the tuning indicator. It is possible that the voltage selector being set incorrectly damaged them. Or maybe the voltage selector had been set to a lower setting to get the radio working again as the valves aged.
Considering the fact that the set was full of Hunts capacitors, most of which were falling to pieces as these tend to do, it was actually working fairly well. However it was still not quite right. The tone control only gave reasonable sound in one position and crackled badly as it was operated (even after applying contact cleaner). Tuning, particularly on VHF, was not as precise as it should be. And the tuning indicator was still not responding.
Since I was going to change these capacitors anyway, I decided to do this now, because I expected this to solve the outstanding problems. There were about fifteen in total, most of which I replaced with dipped polyester types intended for PCB mounting (because they were the cheapest I had).
One capacitor was visible on the top of the VHF
head, so I used one of those beige coloured translucent Wima capacitors which
looked better than a more modern part. I also replaced the small 5uF electrolytic
capacitor in the FM ratio discriminator detector circuit.
While I was replacing the capacitors, I discovered another previous repair that I had missed before. The volume control pot had been replaced. The engineer had cut the tags from the faulty control and soldered them onto the tags of the new pot. This was done neatly, because I only discovered it when unsoldering the connections to change a capacitor.
I also found a capacitor with one end completely disconnected. By reference to the circuit diagram, this was part of a treble-boost circuit around the volume control, which is switched in on the AM bands only. The related resistor was missing. I wondered whether this circuit had been removed due to another switch fault, but there was no sign of discolouration on the switch. I reconnected the circuit as it should have been, knowing that if there were problems I could remove it again.
Unlike many sets, this chassis is easy to work on. The mounting brackets extend upwards at the front and rear forming support brackets that allow the chassis to be placed upside down or stood on either end steadily, with no risk of damaging components.
Further Tests and Repairs
Following the capacitor replacement session, the set worked a lot better. In fact it sounded really good - and this was with the speaker just sat on the bench.
I then returned to the HT switching section of the waveband switch. As with the other section that had been disconnected, I could see no discolouration. I cleaned the switch with some contact cleaner and a toothbrush, then reconnected the circuit as it was supposed to be. It worked fine on all wavebands, and there was no crackling or rustling after a few hours soak testing.
So why had the switches been disconnected? My guess is that the set required repairs for a crackling or rustling sound, and the engineer focused on the likely sources of this fault (tracking on switches) before eventually realising that the problem was the on/off/volume control (possibly the switch section). Once this was replaced the set worked, so the engineer didn't bother to reinstate the original connections. Maybe he had spent enough time on the job already.
The only electrical job remaining was to replace the mains flex. The original cable was a three-core type with the earth connected to chassis, and I fitted a three-core replacement the same way. The original cable clamp was a wedge-shaped sleeve, which is slid over the cable (it's a snug fit) and pressed into a hole in the chassis from the inside. Although it seems a bit uninspiring, it actually worked very well on the new cable. It would however be possible to push the cable into the set and loosen the grip, so I applied a bit of Superglue where the sleeve fits into the hole to give it some extra support.
On the home stretch now! The chassis was originally covered in a thick layer of brown dust. The worst of this had been brushed off before I started, and more had been removed during the electrical repairs, but it was now to time to give it a thorough clean, including all the inaccessible gaps. I used an old toothbrush and a paintbrush to clean every speck of dust I could get to. Small gaps between and underneath components were cleaned with cotton buds, small paintbrushes, folded tissue paper and anything else that would fit.
Underneath the dust the chassis was in very good order. It had originally been plated, and this plating was still sound so there was no rust anywhere. I decided against any further (wet) cleaning because it looked presentable as it was. The HT decoupling resistor mounted on the output transformer consisted of two 3K3 2W carbon resistors in parallel. These were absolutely filthy, and attempts to clean them resulted in the markings being removed while the dirt still remained! They looked a mess so I replaced them with a single 1K8 10W wirewound resistor. This was a grey RS type probably made in the 70s, which didn't look out of place.
The white metal plate behind the tuning scale pointers was anything but white! The paint was badly discoloured in one patch, and looked very dirty. I removed it from the chassis for attention. It wouldn't clean with foam cleaner, and T-Cut did not have any worthwhile effect. I painted the whole plate with aerosol car paint in Ford Sierra Beige (I had a part-tin left over from another job), and left it to dry overnight before refitting.
The valves, which had been removed for safety while cleaning the chassis, were themselves carefully cleaned (taking care not to disturb the markings) before being refitted.
Once the knobs were cleaned with foam cleanser and Brasso, the set was carefully reassembled. Everything went together with no problems.
The only section I did not bother to resurrect was the internal VHF aerial. This consisted of foil strips glued inside the top and sides of the cabinet, most of which was now missing. The piece of cable linking this to the socket on the chassis had been cut short and the plug was missing. This sort of aerial never works very well in my experience anyway, so I decided to stick with an external wire aerial.
The finished set sounds really good. It was clearly an expensive radio when it was originally sold (£26 5s in 1955 according to the Trader sheet), and the original owners must have been very pleased with it. There were many sets that received the new VHF band, but few did it justice as well as this set.
Unfortunately it only tunes up to 101MHz so it won't receive Saga FM. However the first test was on a Sunday afternoon, so I tuned it into BBC Radio 2 and thoroughly enjoyed listening to the music played on Desmond Carrington's All Time Greats and the following programme, while I finished typing up this report.
The tone control is a four-position switch that applies varying degrees of treble cut. I think it sounds best on position "1", maximum treble. Position "2" softens it a little and is still very pleasant. Positions "3" and "4" were probably intended for those listeners at the time who were not used to the bright sound of VHF broadcasts and preferred a more mellow tone. MW and LW are also surprisingly clear without being shrill or sounding off-tune. The treble boost circuit on these bands probably helps here. Even weak stations on MW and LW sound good.
Well at the end of that epic I have turned a pile of rotten junk that was destined for the dustbin into a presentable looking radio that also sounds good. At a cost of...
Six valves (probably £30 if I had to buy them all), fifteen capacitors (about £5), cabinet materials (£20 spent plus lots I already had), etc. Plus probably 40 or 50 hours work. And the result is a set that's worth probably £30 in auction.
No, it doesn't add up. In pure financial terms, radio restoration is rarely worthwhile, particularly if you have to consider the time at an hourly rate. But to us radio collectors and restorers, that's irrelevant.
Restoring sets like this is a useful exercise, because you can learn from your mistakes before tackling something that is actually worth some real money. Probably the main mistake on this set was the choice of varnish. The satin finish and the colour was OK, but the drying speed was too quick.
Some of the repairs are visible, but I think that's unavoidable with the way I tackled it. For example, the join between the original wood and new veneer on the bottom corner of the side is visible. The solution would be to completely veneer the side. If it were a more valuable set, this sort of thing would have to be done differently.
But despite all that, I am very pleased
with the result.